**UPGRADE YOUR BROWSER**

We have detected your current browser version is not the latest one. Xilinx.com uses the latest web technologies to bring you the best online experience possible. Please upgrade to a Xilinx.com supported browser:Chrome, Firefox, Internet Explorer 11, Safari. Thank you!

Turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

- Community Forums
- :
- Forums
- :
- Hardware Development
- :
- AI Engine, DSP IP and Tools
- :
- Reciprocal of fixed point number

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Mute
- Printer Friendly Page

sridar-s

Adventurer

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

03-06-2018 08:11 AM

1,354 Views

Registered:
01-09-2011

1 Solution

Accepted Solutions

hbucher

Scholar

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

03-06-2018 10:17 AM

1,563 Views

Registered:
03-22-2016

@sridar-s The truth is, fixed point numbers are integers. You just use the decimal information to convert from/to a string representation. The rest is all integer algebra.

Fixed point division is integer division, there's not much to do about it.

vitorian.com --- We do this **for fun**. Always give kudos. Accept as solution if your question was answered.

*I will not answer to personal messages - use the forums instead.*

4 Replies

hbucher

Scholar

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

03-06-2018 10:17 AM

1,564 Views

Registered:
03-22-2016

@sridar-s The truth is, fixed point numbers are integers. You just use the decimal information to convert from/to a string representation. The rest is all integer algebra.

Fixed point division is integer division, there's not much to do about it.

andrzejprzybyl

Newbie

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

08-24-2019 01:15 AM

548 Views

Registered:
08-24-2019

Re: Reciprocal of fixed point number

I think you're wrong.

Finding the arithmetic reciprocal is a bit simpler than dividing, because the value of the dividend in the first case is predetermined, i.e. equal to 1.0 (in the interpretation of integers as fixed point values).

It follows that the hardware implementation of the inverse arithmetic determination should be somewhat simpler, i.e. use less hardware resources and determine the result in a smaller number of cycles.

So, in my opinion, THERE IS something important to do here, i.e. adding the appropriate Core IP.

u4223374

Scholar

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

08-24-2019 01:45 AM

540 Views

Registered:
04-26-2015

Re: Reciprocal of fixed point number

**very** good at logic optimization. I strongly suspect that if you just tie the dividend input to a constant value, they'll happily optimize away a lot of the block and produce hardware that looks very much like a purpose-built reciprocal calculator.

drjohnsmith

Teacher

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

08-27-2019 03:20 AM

493 Views

Registered:
07-09-2009

Re: Reciprocal of fixed point number

how big is the number ?

how oftern does it change ?

If its defined at compile / synthesis , then you can do it for free as a constant.

If its say 16 bits or less , then use a Look Up table in a BRAM

If it changes slowly, use a recursive divide, as you would on paper

<== If this was helpful, please feel free to give Kudos, and close if it answers your question ==>